Guns detection using YOLO algorithm

This approach focuses on the technology of object detection, the various and innovation methods used in this field of machine learning in general, focusing on object detection in particular. In which we clarify the fundamental things in this field, which have been developed gradually and become the focus of attention to the development of the world at present. The practical aspect of this research explains the use of the You Only Look Once (YOLO) technique, as this branch of object detection is the most developed and most updated branch used in the last couple of years. Our idea is detecting guns in real-time videos or monitoring cameras, so we need the speed of processing that could process frames per second; this property is available in YOLO as the best technique having the speed of the process. We have created a dataset for a specific weapon (AKM-47) as a test for the technique if the camera can detect weapons in real-time or not. We mentioned that despite the sophistication of this technology and its need for some large specifications such as a high-speed memory like SSD's, a good CPU's and also needs GPU's. But it still is very important to process it, because if it is applied on the ground of its art, it will help many people avoid becoming victims of terrorism and murder.